Skip to main content

Full text of "Natural polymorphisms and unusual mutations in HIV-1 protease with potential antiretroviral resistance: a bioinformatic analysis."

See other formats


Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Bioinformatics 



RESEARCH ARTICLE Open Access 



Natural polymorphisms and unusual mutations in 
HIV-1 protease with potential antiretroviral 
resistance: a bioinformatic analysis 

Carlos Mata-Munguia 1 , Martha Escoto-Delgadillo 2,6 , Blanca Torres-Mendoza 2,7 , Mario Flores-Soto 2,8 , 
Mildred Vazquez-Torres 2 , Francisco Galvez-Gastelum 3 , Arturo Viniegra-Osorio 4 , Marcelo Castillero-Manzano 5 
and Eduardo Vazquez-Vails 2,5 * 



Abstract 

Background: The correlations of genotypic and phenotypic tests with treatment, clinical history and the 
significance of mutations in viruses of HIV-infected patients are used to establish resistance mutations to protease 
inhibitors (Pis). Emerging mutations in human immunodeficiency virus type 1 (HIV-1) protease confer resistance to 
Pis by inducing structural changes at the ligand interaction site. The aim of this study was to establish an in silico 
structural relationship between natural HIV-1 polymorphisms and unusual HIV-1 mutations that confer resistance to 
Pis. 

Results: Protease sequences isolated from 151 Mexican HIV-1 patients that were naive to, or subjected to 
antiretroviral therapy, were examined. We identified 41 unrelated resistance mutations with a prevalence greater 
than 1%. Among these mutations, nine exhibited positive selection, three were natural polymorphisms {L63SA//H) in 
a codon associated with drug resistance, and six were unusual mutations {L5F, D29V, L63R/G, P79L and T91V). The 
D29V mutation, with a prevalence of 1 .32% in the studied population, was only found in patients treated with 
antiretroviral drugs. Using in silico modelling, we observed that D29V formed unstable protease complexes when 
were docked with lopinavir, saquinavir, darunavir, tipranavir, indinavir and atazanavir. 

Conclusions: The structural correlation of natural polymorphisms and unusual mutations with drug resistance is 
useful for the identification of HIV-1 variants with potential resistance to Pis. The D29V mutation likely confers a 
selection advantage in viruses; however, in silico, presence of this mutation results in unstable enzyme/PI complexes, 
that possibly induce resistance to Pis. 

Keywords: Antiretroviral agents, Bioinformatics, Molecular docking simulation, Drug resistance, HIV protease, 
In silico, Polymorphism, Mutations 



Background 

Diversity of viral populations is a result of sophisticated 
recombination, replication and/or selection events that 
induce drug-resistant human immunodeficiency virus 
type 1 (HIV-1) variants. The lack of reverse transcription 
corrections, transitional printing and transversion muta- 
tions, along with viral recombination, has resulted in the 



* Correspondence: eduardo.vazquez@imss.gob.mx 

2 Laboratorio de Inmunodeficiencias y Retrovirus Humanos, Centro de 

Investigation Biomedica de Occidente, CMN0, IMSS, Guadalajara 44340, 

Mexico 

5 UMAE, Hospital de Especialidades, CMNO, IMSS, Guadalajara 44340, Mexico 
Full list of author information is available at the end of the article 

Bio Med Central 



emergence of HIV-1 variants with high resistance to 
pharmacological stressors [1,2]. These variants form 
populations that evade antiretroviral agents, due to 
emerging phenotypic changes within and around the ac- 
tive enzyme site [3]. These mutations, which give rise to 
drug resistance, result in reduced efficacy of highly ac- 
tive antiretroviral therapy (HAART) [4]. Correlations be- 
tween genotypic and phenotypic tests with treatment, 
clinical history, and significance of mutations identified 
in HIV-1 of infected patients are used to determine the 
presence of mutations that confer resistance to protease 
inhibitors (Pis) [1]. 



© 2014 Mata-Munguia et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the 
Creative Commons Attribution License (http://creativecommons.Org/licenses/by/2.0), which permits unrestricted use, 
distribution, and reproduction in any medium, provided the original work is properly credited. 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 2 of 17 



Disruption at interaction sites causes an alteration in af- 
finity between proteins and their inhibitors, and has been 
recognized as a property of drug resistant HIV-1 proteins 
[5,6]. Protein folding simulation models can create Local 
Elementary Structures (LES). These secondary structures 
are stabilized by amino acids that interact with the polypep- 
tide chain [7]. Using the Gromacs software (version 3.0), 
LES were found to form in protease (PR) regions 23-33, 
74-78, and 83-92, and also docked in a folding nucleus [8]. 
Other studies have shown that mutations further from the 
active site can alter the flexibility of HIV-1 PR, inducing 
structural changes that affect the efficacy of most Pis cur- 
rentiy used [9]. Theoretical studies, either alone or in com- 
bination with experimental methods, have pointed to an 
increase in the flexibility of mutant enzymes at various sites, 
including the active site, as a resistance mechanism that 
causes a decrease in the affinity of Pis [10]. Part of the 
cause of such flexibility could be the unusual mutations 
that generally emerge only after "major" and "minor" resist- 
ance mutations have been introduced [11]. Other muta- 
tions that can affect the interaction between PR and Pis are 
natural polymorphisms and unusual mutations in positions 
that confer drug resistance. Although the main mutations 
associated with drug resistance have been characterized 
[12,13], little is known about the influence of natural poly- 
morphisms and unusual mutations with respect to the de- 
velopment of drug resistance. The aim of this study was to 
describe an in silico experiment that showed structural cor- 
relations between natural HIV-1 polymorphisms and un- 
usual HIV-1 mutations in the PR region of HIV-1 pol with 
potential Pis resistance. 

Methods 

Sequence data 

We analysed 151 HIV-1 sequences from Mexican pa- 
tients who had been tested for resistance to antiretro- 
viral drugs between 2005 and 2011 in the Laboratory of 
Immunodeficiencies and Human Retroviruses, Western 
Biomedical Research Center, Mexican Institute of Social 
Security. Sequences were obtained from 22 naive, and 
129 treated patients that were not responsive to drugs. 
Sequences were registered in the GenBank database 
[14], with the following accession numbers: [EU045452- 
EU045489; GU382757-GU382851; GU437199-GU4372 
00; and KC416212-KC416227]. All sequences were ana- 
lysed for the presence or absence of highly mutated se- 
quences using HYPERMUT software (version 2.0) [15]. 
For a reference sequence, we used the subtype B consen- 
sus sequence, which was derived from an alignment of 
subtype B sequences maintained at the Los Alamos HIV 
Sequence Database (LANL), and available from the 
HIV Drug Resistance Database (HIVDB), Stanford Uni- 
versity [16]. 



Phylogenetic analysis 

Nucleotide homology analysis for HIV-1 sequences was 
conducted using the NCBI Genotyping Tool program 
[17]. Subtype determinations were further confirmed by 
phylogenetic analysis performed with the Molecular Evo- 
lution Genetics Analysis (MEGA) software package (ver- 
sion 5.0) [18], which includes the recommended 
reference sequence sets, available from the Los Alamos 
HIV Sequence Database [19]. Prior to all phylogenetic 
analyses, HIV-1 pol sequences were aligned using Clustal 
X (European Bioinformatics Institute, EMBL) [20]. Se- 
quences with 100% homology were excluded from the 
analysis. The nucleotide distance matrix was generated 
using the Kimura two-parameter Neighbour-joining 
method [21]. The statistical robustness of the generated 
trees was verified by bootstrap analysis of 1000 
replicates. 

Detection of multidrug resistance phenotypes in HIV-1 
protease 

The genetic changes associated with drug resistance in viral 
sequences were established according to HIVdb algorithm 
version 6.0.9 (http://hivdb.stanford.edu) [22]. The interpret- 
ation of drug resistance was performed at various levels of 
susceptibility for the following USA Food and Drug 
Administration (FDA)-approved Pis: atazanavir (ATV); dar- 
unavir (DRV); amprenavir (APV); indinavir (IDV); lopinavir 
(LPV); saquinavir (SQV); tipranavir (TPV); nelfinavir 
(NFV);and ritonavir (RTV). The resistance mutations were 
classified as major or minor according to HIVdb criteria, or 
as natural polymorphisms or unusual mutations if they 
were not associated with resistance [16]. The prevalence {p) 
for each mutation in the protease region of pol was quanti- 
tatively determined as the frequency of the mutation (M) 
among total sequences evaluated for each position (TV), p = 
M/N, using Microsoft Excel 2010. The genetic variation 
was calculated as the total number of mutations at a nu- 
cleotide position divided by the number of evaluated se- 
quences. The Phenotypic Variation (PV) was defined as the 
percentage (%) of amino acid substitutions for each position 
relative to the consensus sequence. For each region, the PV 
was classified as follows: conserved, <1%; semi-conserved, 1 
to <5%; variable, 5 to <10%; and highly variable, >10%. 
Values found below the 15th percentile and above the 75th 
percentile were not considered. Phenotypic mutations with 
a prevalence of >1.0% among 151 amino acid sequences 
were compared for each PI against the IAS-USA drug re- 
sistance mutations list [12]. The structural conservation of 
PR was defined in a complementary way to that of PV. 

Analysis of selective pressure 

The selective pressure and reconstruction of the ances- 
tral state for each PR codon was determined using a 
maximum likelihood (ML) substitution model and the 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 3 of 1 7 



HyPhy algorithm, included in MEGA5 package [23,24]. 
The synonymous site divergence (dS) and nonsynon- 
ymous site divergence (dN) per branch was estimated 
using the Muse-Gaut codon model [25]. The values of 
the ML model were estimated from the topology of the 
phylogenetic tree. The probability of rejecting the 
hypothesis of neutral evolution was significant with 
p < 0.05. The standardized values of dN-dS were ob- 
tained by the total number of substitutions in the tree 
(calculated as substitutions expected by site). To distin- 
guish between drug pressure and immune system pres- 
sure, results were compared using the HIV positive 
selection mutation database [26]. 

Molecular modelling 

Once the natural polymorphisms and unusual mutation co- 
dons with positive selection (dN/dS > 1) and prevalence 
>1% were obtained, homology modelling was used to pre- 
dict changes in the PR structure. Homology modelling of 
natural polymorphisms and unusual mutations followed 
these steps: (i) template selection; (ii) structural alignment; 
(iii) model construction; and (iv) refinement [27]. To select 
the template, HIV-1 protease X-ray crystal structure 
FASTA sequences available from the Protein Data Bank 
(PDB) [28] and the HIV-1 subtype B consensus sequence 
available from the HIVdb were aligned using ClustalX [20]. 
The PR sequence exhibiting greatest identity with HIV-1 
subtype B consensus (wild-type template) was chosen as 
the template for modelling mutant proteases (PRs). The re- 
sistant PRs used for reference were modelled with each 
major PI resistance mutation. Every mutant protein was 
modelled using a Swiss -Model workspace, which showed 
the identity (%). The expected alignment value with the 
template sequence (E) and the Qualitative Model Energy 
Analysis (QMEAN4), which estimates the absolute quality 
model, ranged from 0-1 [28,29]. 

Estimation of the free energy of binding 

Using the Autodock/Vina application on a LINUX plat- 
form, which had the PyMOL (version 1.4.1) molecular 
graphics system installed, we estimated the free energy 
of binding of the complex between mutant PR structures 
and Pis [30]. Rectangular boxes were used to define the 
binding sites and these were adjusted by providing spe- 
cific coordinates of active PR sites before each docking. 

Receptor and ligand representations in the Protein Data 
Bank, Partial Charge & Atom Type formats (pdbqt) con- 
taining atomic charges, atom type definitions and topo- 
logical information, were produced using Autodock/Vina 
[30] . To determine if the differences caused by natural poly- 
morphisms and unusual mutations had any effect on the 
free energy of binding of Pis, the free energy values ob- 
tained for the resistant protease/ligand complexes were 
compared. Natural polymorphisms or unusual mutations 



with lower or equal affinity to Pis compared with reference 
proteins containing drug resistance mutations indicate 
positive resistance. Higher affinity was considered to favour 
susceptibility of the HIV-1 variant to Pis. The coupled pro- 
teases included the wild-type PR [PDB: 1GNO], PRs with 
major drug-resistance mutations and PRs with natural poly- 
morphisms or unusual mutations at codons having positive 
selection. 

Measurement of distances between protease residues 
and Pis 

To evaluate the natural polymorphism and unusual mu- 
tation atoms that affect the affinity of Pis, we measured 
the distances (A) between the amino acid residue C„- 
atoms implicated in drug resistance, and the closest het- 
eroatoms of the Pis. Complexes that showed signs of 
free energy of binding were analysed, suggesting in- 
creased drug resistance because of the presence of nat- 
ural polymorphisms and unusual mutations. Distances 
were compared with those obtained for the same pair of 
atoms in the wild-type and resistant PR structures avail- 
able from the PDB [28]. All interatomic distances were 
measured with PyMOL (version 1.4.1) [31]. 

Results and discussion 

Genetic relationships of HIV-1 variants isolated from 
Mexican patients 

Phylogenetic analysis of the 151 HIV-1 protease fragment 
nucleotide sequences was conducted using a Neighbour- 
joining tree. Phylogenetic relationships were grouped into 
the internal nodes of the tree, using subtype B reference se- 
quences [GenBank: U63632 and U21135]. The HIV-1 vari- 
ants isolated from Mexican patients, and confirmed by 
analysis with the NCBI Genotyping Tool, were subtype B. 
This result is consistent with other molecular epidemiology 
studies of Mexican HIV-1 patients, with or without anti- 
retroviral intervention, where subtype B prevails [32,33]. 

Drug resistant phenotypes and genotypes of HIV-1 
protease 

Non-synonymous genetic changes largely contribute to 
phenotypic changes [34]. Because of degeneracy in the gen- 
etic code, transcription and translation errors during the 
viral replication cycle, along with functional, structural, 
pharmacological and immunological selection pressures, 
there is no absolute mathematical relationship between gen- 
etic and phenotypic variations [35]. Variations in the pri- 
mary structure of the 151 PR sequences are presented in 
Figure 1. Of the 36 codons associated with major or minor 
resistance, 19 showed PV <5% (Lll, L24, D30, V32, L33, 
E34, K43, 147, G48, 150, F53, D60, G73, T74, L76, N83, 185, 
N88 and L89). Among these codons, G48, 150 and F53 were 
present in conserved regions, and L24, D30, L76 and N88 in 
semi-conserved regions. Six codons (G16, K20, 154, Q58, 



Mata-Mungui'a et al. BMC Bioinformatics 201 4, 1 5:72 Page 4 of 1 7 

http://www.biomedcentral.com/1471-2105/15/72 



a Genetic Variation % 

RNA 

Amino Acid 
b Phenotypic Variation °/c 
Region (PV) 



0.0 0.0 00 

C C T 

p 

0 



0.0 0.0 0.0 

C A G 

Q 
0 



0.0 0 0 0 0 

A T C 



00 0.0 0.0 

A C U 

T 

0.15 



2.10 0 00 0 78 

c u u 

L 
1.67 



5.00 


1.14 


1.52 


0.00 


1.12 


0.75 


G 


G 


c 


A 


A 


C 


w 






Q 






S.2 






1.89 







G 

R 

1.12 



c c clul 

0.36 | 17.83 



Drug Resistance 




1 




2 




3 






4 






5 






6 






7 






8 






9 






10 






0.00 


0.68 


5.03 


6.71 


3.36 


2.70 




0.33 


2.65 


1.66 


9.93 


8.61 


8.94 


0.33 


0.66 


0.99 


4.64 


4,97 


0.00 


3.31 


9 93 


1.32 


0.66 






1.99 


4.97 


1.32 


7.95 


8.61 




G 


u 


c 


A 


c 


A 




U 


A 


A 


A 


G 


A 


u 


A 


G 


G 


G 


G 


G 


G 


c 


A 


7 




u 


A 


A 


A 


G 






v 






T 


















I 






G 






G 






Q 






L 






K 








0.68 






9.4 


















9 27 






563 






3.31 






3£4 






7.95 






7.95 






































8.04 
































11 






12 






13 






14 






15 






16 






17 






18 






19 






20 




8 G8netic Variation % 


0.33 


o oo 


2.32 


0.66 


0.66 


7-95 


1.66 


0.66 




3.97 


0 66 


3.31 


0.00 


0.00 


4.64 


0.00 


0.99 


0.66 


0.66 


0.00 


3.31 


0.66 


0.99 


0.66 


0.00 


0.99 


1.32 


1.32 


0.00 


2.65 


RNA 


G 


A 


A 


G 


C 


u 


c 


u 




u 


u 


A 


G 


A 


u 


A 


c 


A 


G 


G 


A 


G 


c 


A 


G 


A 


u 


G 


A 


u 


Amino Acid 
b Phenotypic Variation "A 




E 

0 33 






A 

1.32 






L 
1 32 






L 

1 66 






0 
0 






T 

0.99 






G 

0.66 






A 

0 99 






D 
2.65 






D 
1.32 




Region (PV) 


■ — 








































1.24 


















Drug Resistance 




21 






22 






23 






24 






25 






26 






27 






28 






29 






30 






0.66 


0.66 


1.66 


1.66 


0.00 


7.95 


3.64 


0.00 


1.66 


0.33 


0.33 


1.99 








0.99 


0.00 




2351 


1987 


7.62 


2.65 


0.00 




4.30 


1.32 


0.00 


0.00 


0.00 






A 


c 


A 


G 


U 


A 


U 


U 


A 


G 


A 


A 




n 




A 


u 


G 


A 


A 


u 


u 


U 




c 


c 


A 


G 


G 








T 






V 






L 






E 










M 






N 






L 






p 






G 








1.16 






1.99 






2.32 






0 66 












19.54 






40.07 




1.66 






4.97 






0 








































31 






32 






33 






34 






35 






36 




37 




38 






39 






40 




a Genetic Variation % 


0.00 




7.95 


0 33 


0.33 


0.33 


0.33 


3.64 




0.66 


0.66 


2.32 


0.66 


0.66 


0.66 


2.98 


0.00 


7.62 


0.99 


0.66 


0.33 


0.00 


0.66 


0.66 


0.00 


0.00 


1.32 


0.00 


0.66 


0.00 


RNA 


A 




A 


u 


G 


G 


A 


A 




c 


c 


A 


A 


A 


A 


A 


U 


G 


A 


u 


A 


G 


G 


G 


G 


G 


A 


A 


u 


U 


Amino Acid 
b PhenotypicVariation % 




R 
19 






w 

0 83 






K 
3.97 






p 

1.32 




K 

1.32 




M 
10.26 


2.65 




G 

0.66 






G 
0 






I 

0.66 




Region (PV) 
Drug Resistance 


2.39 






41 






42 






43 






44 






45 






46 






47 






48 






49 












0.00 


0.00 


1.32 


0.00 


0.66 


4.30 


0.66 


0.66 


0.00 


5.30 


0.00 


3.31 


0.00 


0 00 




0.99 


0.00 


6.62 








3.64 


0.66 


894 


0.00 


2.32 


0 99 


0.33 


0.33 


6.95 




G 


G 


A 


G 


G 


u 


u 


U 


U 


A 


u 


c 


A 


A 




G 


U 


A 








c 


A 


G 


U 


A 


U 


G 


A 


u 






G 






G 






F 






I 






K 






V 












a 






Y 






D 








0 






033 






0 66 






5.3 






0 


















5 96 






3 31 






4 64 




























1 




















Figure 1 (See legend on next page.) 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 5 of 1 7 



(See figure on previous page.) 

Figure 1 Genetic and phenotypic representation of primary structure variation within the HIV-1 protease consensus. Codons 13 are 
shown in grey and were not included in our analysis. Conserved regions are shown in yellow, semi-conserved regions in ochre, variable regions 
in orange and highly variable regions in red. The major (dark green) and minor (light green) resistance mutations are indicated for each codon. 
a Genetic variation = total number of mutations at the nucleotide position/number of sequences evaluated. b Phenotypic variation = total number 
of mutations at the amino acid position/number of sequences evaluated. 



H69 and 184) displayed PV of 5-10%. Of these, only codons 
154 and Q58 were located in conserved and semi-conserved 
regions, respectively. The remaining 11 codons associ- 
ated with drug resistance (L10, M36, M46, 162, L63, 
164, A71, V77, V82, L90 and 193) had variation >10%, 
with L10 and 193 present in conserved regions, and 
M46, V77 and L90 in semi-conserved regions. There 
was also a variable PV of 5-10%, in the codons neigh- 
bouring the drug resistance positions (T12, 115, L19, 
G68, and K70), with codons E35, N37, R41, R57, and 
172 highly variable. Figure 2 illustrates the mutations 
with prevalence >1% found in the protease region of 



HIV-1 that were present in the 151 PR sequences ex- 
amined. According to the IAS -USA, the mutations as- 
sociated with drug resistance, with a p >10%, were 
L10I, M36I, 162V, L63P, 164V, A71V/T, V77I, L90M, 
and I93L [12,36,37]. 

Structural studies of PRs have reported a slight widen- 
ing of the active site due to mutations associated with 
drug resistance for the majority of Pis [9,10,38]. How- 
ever for other inhibitors, such as IDV which is charac- 
terized by three aromatic rings, structural changes are 
caused by mutations at the active site and adjacent posi- 
tions [39]. 



0> 
o 
c 

re 
> 



| Mutations associated with drug resistance 
^Mutations not associated with drug resistance 



h 



5 = S2^SKS5SSSt|2S£Sgg|5gg3cgfeKtg5BSR 



70.8 



Ilh 



LI 



Mutations 

Figure 2 Prevalence of mutations within HIV-1 PR pol. Red bars represent mutations associated with drug resistance, and the green bars 
represent natural polymorphisms and unusual mutations not associated with drug resistance. 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 6 of 1 7 



Table 1 Polymorphisms or unusual mutations (p> 1%) weakly associated with PI resistance in HIV-1 protease from 
treated and naive individuals according to the HIVdb 



Mutation 


p (%) 


PV 

(%) 


Region 
(PV) 


Association with drug resistance 


Classification 


W6R 


2.34 


8.20 


C (0.98) 


Found in indinavir-resistant PR [44] 


UM 


T12A/I 


1.34/ 
1.34 


9.40 


V (8.04) 


T12A decreased in patients treated with Pis. [45] 1121 appears in cell culture in the presence 

of SQV [46] 


NP/NP 


113V 


17.33 


18.33 


V (8.04) 


Found in isolates from patients treated with NFV [47] 


NP 


115V 


8.28 


9.27 


V (8.04) 


Associated with reduced virological response to RTV + SQV therapy [48] 


NP 


E35D 


18.21 


18.54 


HV(26.05) 


Associated with reduced in vivo virological responses to RTV/AMP [49] 


NP 


N37D/E 


9.27/ 
5.96 


40.07 


HV(26.05) 


N37D appears together with N37E in patient treated with LPV + RTV [50] 


NP/NP 


R41K 


19.00 


19.00 


SC(2.39) 


Associated with reduced in vivo virological responses to RTV + APV in Pis experienced 

patients [49] 


NP 


R57K 


17.88 


18.38 




Relatively frequent in patients failing treatment with RTV + SQV [51] 


NP 


L63A 


3.48 


84.77 


HV(22.02) 


L63A frequent polymorphism but significantly associated with the antiretroviral treatment 

[39,52] 


NP 


H69Y 


3.15 


7.62 


V(7.56) 


Appears in viruses selected with LPV [53] 


NP 


K70E 


3.31 


9.11 


V(7.56) 


K70E appears in virus selected in cell culture with DRV [54] 


NP 


172R 


1.32 


24.50 


HV(27.48) 


Associated with viral rebound during therapy with LPV+ RTV [50] 


UM 



p, prevalence; PV, phenotypic variation; C, conserved; SC, semi-conserved; V; variable; HV, highly variable; APV, amprenavir; ATV, atazanavir; DRV, darunavir; IDV, 
indinavir; RTV, ritonavir; SQV, saquinavir; Pis, protease inhibitors; NP, natural polymorphisms; UM, unusual mutations. 
*Values below the 15th percentile or above the 75th percentile were not considered. 



Prevalence of natural polymorphisms and unusual 
mutations in PRs without established drug resistance 

Table 1 shows the natural polymorphisms or unusual mu- 
tations with a p >1% that were found in the PR sequences 
of HIV-1 isolated from the Mexican patients. These are 
weakly associated with PI resistance, but are not included 
in the IAS-USA guides or the HIVdb as accessory or 
minor mutations [16,40,41]. Of the 14 mutations, only 
L63A and H69Y were found in drug resistance positions, 
and T12A/I, I 15V, E3SD, N37D/E, R57K, K70E and 172V 
were contiguous to positions associated with resistance. 
Overall, these mutations have little effect on drug suscep- 
tibility; however, a phenotypic change in any of them 
could have relevance to the affinity to one or more Pis 
[6,42]. These mutations, in combination with resistance 
mutations, might have an effect on the dynamics of the 
evolution of cross-resistance [43]. 

The I13V (17.33%), E35D (18.21%), R41K (19%) and R57K 
(17.88%) mutations had a p > 10% and were located in poly- 
morphic positions observed in non-B subtypes [35,55,56]. In 
the HIVdb, W6R and I72R are unusual mutations with a 
frequency <0.05% that only emerge after multiple major and 
minor resistance mutations [57]. Table 2 shows 41 muta- 
tions with ap >1% that have not been associated with resist- 
ance, 25 are natural polymorphisms and the remaining 16 
were unusual mutations. According to phenotypic conserva- 
tion analysis, the L5F and Q7E mutations were within the 
conserved regions, while D29V, P39S, K43R, Q61E, E65D, 
C67F, P79L, T91V and Q92G/K were within semi-conserved 
regions. The T12P/S, K14R, G17D/E, Q18H, L19I/V/T, 



G68E, H69Q and K70R/T/I mutations were within the vari- 
able regions, and N37S/T/C/H/I, L63S/V/R/G/H, I72V/T/E/ 
M and 193F were in highly variable regions. 

Among the codons presented in the Table 2, the muta- 
tions in positions K43, L63, H69 and 193 were located in 
sites associated with minor resistance, but the distance be- 
tween its localization and the enzyme's active site reduces 
the possibility of the structure contributing to drug resist- 
ance. All the described mutations could be due to random 
transcriptional errors, or positive selection from drug and/ 
or immunological stressors [37,58]. Generally, natural 
polymorphisms occur in remote regions away from the ac- 
tive site, and form domains that define the shape of the 
homodimer. However, unusual mutations are found in po- 
sitions associated with drug resistance and possibly gener- 
ate allosteric changes in the binding site that favour 
enzymatic function, or decrease the affinity with certain 
Pis [59]. Therefore, the study of such structural changes 
produced by these emerging mutations may help in deter- 
mining the new effects of Pis with different affinities. 

Figure 3 shows PR tertiary structure positions that are: 
not associated with PI resistance; weakly associated with 
PI resistance; associated with PI resistance. We have also 
presented the locations of natural polymorphisms and un- 
usual mutations (Figure 3). The codons T12, N37, L63, 
H69, K70 and 172 include mutations weakly associated 
with PI resistance (T12A/I, N37D/E, L63A, H69Y, K70E, 
and I72R), and mutations lacking evidence of PI resistance 
(T12P/S, N37S/T/C/H/I, L63S/V/R/G/H, H69Q, K70R/T/ 
I, and I72V/T/E/M). 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 7 of 17 



Table 2 Natural polymorphisms and unusual mutations of HIV-1 protease (p> 1%) without evidence of resistance to 
Pis 



Mutation 


P (%) 


PV (%) 


Region (PV) 


Classification 


L5F 


1.67 


1.67 


C (0.95) 


UM 


Q7E 


1.52 


1.89 


C (0.95) 


UM 


T12P/S 


4.03/1 .34 


9.40 


V (8.04) 


NP/NP 


K14R 


9.60 


11.26 


V (9.97) 


NP 


G17D/E 


1 .99/1 .32 


3.31 


V (9.97) 


UM/UM 


Q18H 


1.32 


3.64 


V (9.97) 


NP 


U9W/T 


4.64/1 .32/ 1 .32 


7.95 


V (9.97) 


NP/NP/NP 


D29V 


1.32 


2.65 


SCO .24) 


UM 


N37S/T/C/H/I 


14.4/2.81/1.99/ 1.66/1.32 


40.07 


HV(26.05) 


NP/NP/NP/NP/UM 


P39S 


2.98 


4.97 


SC(2.09) 


NP 


K43R 


3.64 


3.97 


SC(2.09) 


NP 


Q61E 


2.65 


3.97 


SC(4.47) 


NP 


L63SA//R/G/H 


1.99/149/1.99/1.32/1.32 


84.77 


HV(22.02) 


NP/NP/UM/UM/NP 


E65D 


2.0 


2.67 


SC(2.0) 


NP 


C67F 


2.0 


3.33 


SC(2.0) 


NP 


G68E 


4.30 


5.96 


V(7.56) 


UM 


H69Q 


1.99 


7.62 


V(7.56) 


NP 


K70R/T/I 


1.99/1.32/1.16 


9.11 


V(7.56) 


NP/UM/NP 


I72V/T/E/M 


1 1 .26/6.95/2.32/1 .32 


24.50 


HV(27.48) 


NP/NP/NP/UM 


P79L 


1.32 


2.48 


SCO .53) 


UM 


T91V 


3.33 


3.33 


SC(2.15) 


UM 


Q92G/K 


2.03/2.03 


4.05 


SC(2.15) 


UM/UM 


I93F 


1.35 


47.97 


HV(47.63) 


UM 



p, prevalence; PV, phenotypic variation; C, Conserved; SC, semi-conserved; V, variable; HV, highly variable; NP, natural polymorphisms; UM, unusual mutations. 



The D29V and P79L mutations are located near the 
active site of the protease, and therefore possibly con- 
tribute to the generation of PI resistance. It is of interest 
to evaluate these unusual mutations in silico, and estab- 
lish their association with resistance to Pis. 

Phenotypic conservation of HIV-1 protease 

Figure 4 shows the conserved, semi-conserved, variable and 
highly variable regions of PRs according to PV. Mutations 
were clustered into 15 regions, for amino acids 4-99 of the 
protease. For average PV calculation, when the asymmetry 
in the distribution was greater than 1.4 between the 15th 
and 75th percentiles, the residues were not considered. We 
found three conserved, three variable, three highly variable 
and six semi-conserved regions for each chain. The posi- 
tions excluded from the PV calculated for each region were 
W6, L10, 113, K14, G17, Q18, E35, G40, R41, M46, 154, 
V56, R57, 163, V77, N83, L90, Q92, 193 and K97. The PV in 
these codons had very different values from those presented 
by the codons in their respective regions. According to our 
model of protease conservation, the LES formed by frag- 
ments 23-33 and 74-78 were in semi-conserved regions 



(E21-L34 and G73-P81, except for V77). The LES formed 
by the 83-92 fragment involved two codons with variable 
PV, 184 (6.29%) and L90 (12.33%), and two codons with 
semi-conserved PV, T91 (3.33%) and Q92 (4.05%) [8,60]. 
Codon 90 contained a drug resistance mutation {L90M) 
common for most Pis, with the exception of DRV and TPV, 
while T91 and Q92 contained the T91V, Q92G, and Q92K 
mutations, which are classified in the literature as unusual 
mutations. The prevalence of the L90M, T91V, Q92G, and 
Q92K mutations was 12.0, 3.33, 2.03 and 2.03%, respectively. 
Although the effectiveness and specificity of PR proteolytic 
activity is determined by its active site (amino acids 25-29), 
these characteristics are influenced by mutations in neigh- 
bouring structures, which mainly affect intramolecular inter- 
actions with the active site [5,38,42,61]. Contiguous regions 
and the active site have a semi-conserved state, with a PV of 
1.2%. It has been shown that active sites with poor capacity 
to carry out structural changes help adjust the specificity of 
natural substrates without losing proteolytic effectiveness 
[45]. A study that identified the minimal conserved structure 
of HIV-1 PR, in the presence or absence of drug stress, 
showed that most of the PV is a product of pharmacological 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 8 of 17 




Codons without evidence 
of resistance to Pis 



Codons weakly associated 
with resistance to Pis 



Codons associated with 
resistance to Pis 



Figure 3 Codons with natural polymorphisms and unusual mutations in the HIV-1 PR tertiary structure. Codons in the PR that were not 
associated with PI resistance (cyan), weakly associated with PI resistance (yellow), and associated with PI resistance (red). 



stress [62]. In contrast, the peripheral structural regions have 
a relatively high PV (for variable and highly variable regions) 
courtesy of negative selection, and to a lesser extent through 
resistance of HIV-1 to immune stress [63,64]. 

Selective pressure in the pr fragment of HIV-1 pol 

Antiretroviral treatment can exert strong selective 
pressures within pol, which transcribes PR, reverse 



transcriptase and integrase [62,65]. We have presented 
the selection pressure results for 10 codons with natural 
polymorphisms and unusual mutations (Table 3). Ac- 
cording to these results, codons 5, 29, 63, 79, 91 and 93 
represent positive pressure (dN-dS > 1) through the ML 
substitution model using the HyPhy algorithm. When 
these results are compared with the data available in the 
UCLA HIV Positive Selection Mutation Database, only 




Variability range 



<1% 


Conserved 


1 to <5% 


Semi-conserved 





Variability Region 


Position 
Cluster 


Mean 
phenotypic 
variation 


Positions not 
considered 


C 


1 


T4to P11 


0.98 


W6, L10 


v 


2 


L12toG20 


8.04 


113, K14, G17, Q18 


SC 


3 


E21 to L34 


1.24 




SC 


5 


L38 to I47 


2.39 


G40, R41 , M46 


c 


6 


G48 to V56 


0.37 


I54, V56 


SC 


7 


R57 to Q61 


4.47 


R57 


SC 


9 


E65 to C67 


2.00 




V 


10 


G68 to K70 


7.56 




SC 


12 


G73to P81 


1.53 


V77 


v 


13 


V82 to 184 


8.45 


N83 


SC 


14 


185 to Q92 


1.83 


L90, Q92 


c 


15 


193 to F99 


0.44 


I93, K97 



Figure 4 Phenotypic conservation of HIV-1 PR isolated from Mexican patients. A consensus sequence was obtained from 151 individual 
sequences. Regions are shown from red to yellow in proportion to their phenotypic variation (PV,%). Positions with natural polymorphisms and 
unusual mutations at drug resistance codons are shown in the protease model. Mutations were clustered into 15 regions between codons 4-9S 
of the protease. 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 9 of 1 7 



Table 3 Selection pressure for codons with unusual mutations and natural polymorphisms 


Codon 


Triplet 


PV (%) 


dN-dS 


dN-dS (N) 


P value* 


Mutations 


5 


CTT 


1.67 


2.29 


0.48 


0.285 


L5F 


6 


TGG 


8.20 


-5.97 


-1.26 


0.996 


W6R 


7 


CAA 


1.89 


-15.57 


-3.29 


1.000 


Q7E 


29 


GAT 


2.65 


20.20 


4.26 


0.003 


D29V 


63 


CCC 


84.77 


2.01 


0.42 


0.196 


L63A/R/S/V/G/H 


70 


AAA 


9.11 


-1.51 


-0.32 


0.964 


K70I 


79 


CGT 


2.48 


4.44 


0.94 


0.190 


P79L 


91 


ACT 


3.33 


5.01 


1.12 


0.086 


T91V 


92 


CAG 


4.05 


-15.62 


-3.46 


0.990 


Q92G;Q92K 


93 


ATT 


47.97 


28.64 


6.34 


1.000 


I93F 



*P<0.05 was considered statistically significant for positive pressure. 

PV, phenotypic variation; dN, number of non-synonymous substitutions per site (n/N); dS, number of synonymous substitutions per site (s/S); N, normalized; 
P, probability. 



codons 63 and 93 were consistent with positive selection 
under immunological and/or pharmacological stressors 
[26]. The difference in selection pressure for codons 5, 
29, 79 and 91 could be due to variability in the antiretro- 
viral regimen sequences administered to Mexican pa- 
tients. In addition to positive selection, the aligned sites 
often evolve at different rates because of other biological 
factors that include site-specific mutation rates and 
functional constraints of amino acid substitutions [66]. 

The codons that were not associated with resistance due 
to pharmacological stress, and had PV >2% were D29 
(2.65%) and P79 (2.48%). These were located near the ac- 
tive site of the enzyme; T91 (3.33%) was also found to be 
necessary for the establishment of the PR dimer. Codons 
associated with resistance due to pharmacological stress 
and PV >2% were 147 (2.65%), V82 (10.6%) and 184 
(6.29%). Only one of these sequences belonged to a naive 
individual, with a mutation at V82; the remaining 
sequences were from treated individuals (three were 
treated with reverse transcriptase inhibitors only, the re- 
mainder were given reverse transcriptase inhibitors and at 
least one PI). 

Structural prediction of mutant HI V-1 PRs 

The molecular structure of all mutant HIV-1 PRs was 
predicted by comparative homology modelling using the 
wild-type HIV-1 PR as a template [PDB: 1GNO]. This 
structure had higher sequence identity compared with 
the HIV-1 subtype B consensus PR sequence available 
from the HIVdb. Additional file 1: Table SI shows the % 
identity, the expected value of the alignment with the 
template sequence (E), and the score for the absolute 
quality of the models. We modelled the proteins with 
unusual mutations (L5F, D29V, L63G, L63R, P79L and 
T91V), natural polymorphisms (L63H an&L63S), and 
drug-resistant mutant PRs with single mutations or pat- 
terns of mutations {D30N, V32I, M36I, M46I, 147V, 



G48V, I50V, I50L, I54M, QS8E, T74P, L76V, V82A, V82L, 
N83D, N88S, 184V, and L90M). 

The model's accuracy was increased because of the iden- 
tity between the mutant and template sequences; therefore, 
we concluded that the model was suitable for all structures. 
The low E values obtained from the modelled proteins indi- 
cate template identification, and adequate target template 
alignment [27]. The reliability of the predicted structures 
with natural polymorphisms and unusual mutations in drug 
resistance positions ranged 0.87-0.89, while positions for 
major mutation proteins ranged 0.83-0.91. The lower 
QMEAN4 values correlated with mutants containing pat- 
terns of resistance, as a result of the reduced identity of 
these proteins with respect to the template structure [67]. 
The QMEAN4 values were acceptable for all the modelled 
structures. Figure 5 shows the overlapping structures of the 
wild-type PR [PDB: 1GNO] and the D29V mutant, with 
high similarity between both structures, as well as a differ- 
ence in the location of the mutation site (position 29). 

D29 plays a crucial role for the folding of retroviral PRs 
[38,68]. Using crystallography, it has been shown that D29 
forms hydrogen bonds with R87, which partially constitutes 
the highly conserved triad, G86 - R87 - D88 [62]. The loss 
of this specific interaction between a - helix 1 (residues 87- 
91) and D29 destabilizes the dimer interface [69]. The PR 
structures of related viruses such as HIV - 2, equine infec- 
tious anaemia virus (EIAV), feline immunodeficiency virus 
(FIV), rous sarcoma virus (RSV) and simian immunodefi- 
ciency virus (SIV), also demonstrate a proximity between 
side chains R87 and D29 [70]. 

The in silico modelling of mutant proteins generated 
structures very similar to those obtained by X-ray crys- 
tallography. The structures with natural polymorphisms, 
unusual mutations in drug resistance positions, and drug 
resistance mutations obtained by comparative homology 
modelling, were appropriate for molecular docking with 
their respective inhibitors. 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 10 of 17 




D29V mutant HIV-1 
protease 



WT HIV-1 protease 
(PDB: 1NGO) 



Figure 5 Wild-type and D29V mutant protease structures. The structure of the wild-type HIV-1 protease (WT) was obtained by X-ray crystallography. 
[PDB: 1 GNO] (blue), and the mutant protease (green) can be clearly seen, with the red structures corresponding to the oxygen atoms of D29. 



Natural polymorphisms and unusual mutations in PRs 
and their effects on the free energy of binding by Pis 

We have presented the free energy of binding (kcal/mol), as 
well as the average value of the five lowest energy conforma- 
tions for the complexes formed by PRs and the main Pis 
(Table 4). The wild- type PR had the lowest free energy of 
binding for all Pis, except for IDV, compared with mutant 
PRs containing major and multiple drug resistance muta- 
tions. The magnitudes corresponding to the minor values of 
the free energy of binding to the reference protein were: 
wild-type protease- 1GNO < major drug-resistant mutant pro- 
teases < multiple drug-resistant mutant proteases. The Pis 
had the greatest degree of affinity for PR 1GNO, consistent 
with the wild-type PR, whereas reduced affinity for mutant 
PRs was proportional to the number of mutations [46]. 

Among the Pis, IDV demonstrated a higher affinity for 
mutant proteins than PR [PDB: 1GNO]. Additionally, a 
study that correlated the in vivo genetic resistance of HIV-1 
to IDV indicated that the development of resistance occurs 
through the combined effects of several mutations, which 
do not confer a measurable degree of resistance when they 
occur on their own [39]. For the other Pis, significant viral 
resistance has been shown to be a result of the appearance 
of one or two substitutions in drug-resistance positions 
[40,71]. 

The difference between affinities of complexes formed 
by wild-type and drug-resistant PRs indicates some contri- 
bution of phenotypic changes towards PI resistance 
[72,73]. The complexes with the largest differences in- 
volved ATV and DRV, both with a difference of -1.2 kcal/ 
mol. This indicates high susceptibility of both compounds 



to drug resistance mutations. Lower differences were ob- 
served (between -0.7 and -0.3 kcal/mol) for other com- 
plexes, indicating these drug resistance mutations have a 
minor or supplementary effect [72,73] . 

We obtained a positive value when we calculated the dif- 
ference of the free energy of binding between the wild- 
type-IDV complex and the drug-resistant mutant-IDV 
complex. This is consistent with a high genetic barrier to 
resistance for IDV, which has lower susceptibility to drug- 
resistance mutations compared with other Pis [39]. When 
comparing the free energy of binding between the com- 
plexes with drug resistance mutations versus natural poly- 
morphisms and unusual mutation complexes, resistance to 
ATV, LPV, NFV, and TPV was always observed. The PRs 
with L5F, D29V, L63G, L63H, L63R, L63S, and P79L muta- 
tions had lower or equal free energy of binding to ATV, 
LPV, NFV and TPV, than those with wild-type or drug- 
resistant PRs. 

The complex formed by the D29V mutant showed con- 
siderable differences between the distance of the V29 and 
D30 C„ -atoms and the heteroatoms closest to the Pis 
(Table 4). This is probably because of the absence of the 
Cp carboxyl group in the valine compared with the wild- 
type D29. The electrostatic interactions exercised by the 
D29 carboxyl oxygens provide stronger affinity to Pis in 
the active site, resulting in greater affinity compared with 
the V29 mutant [5,74,75]. The absence of V29 carboxyl 
oxygens decreases the level of interactions, thus decreas- 
ing affinity. Such differences can be observed when meas- 
uring the distance between the functional groups of the 
wild-type, resistant and D29V mutant PRs docked to DRV 



Table 4 Free energy of binding for protease-PI complexes obtained in silico 



Inhibitor IGNO-Ligand Mutant protease with major drug resistance mutation kcal/mol Multiple mutant 
kcal/mol kcal/mol 



Mutant protease with emergent mutations L5F/D29V/L63G/L63H/ 
L63R/L63S/P79L/T91 V kcal/mol 



Amprenavir -8,5(-8,4) -8,4(-7,9)//50l/-8,2(-8,12)//84l/ 

Atazanavir -8,3(-8,1) -7,3(-6,94)//50Z. -8,5(-8,14)//84V-8,2(-7,98)//V885 

Darunavir -9,3(-8,96) -9(-8,54)//47V -9,3(-8,86)//50V -9,4(-8,84)//54/W -9,1 (-8,86)/ 184V 

Indinavir -1 0,4(-1 0,02) -10,8(-10,5)/M46/-10,8(10,4)/V82/1 -10,5(-10,28)//84V 

Lopinavir -10,3(-9,84) -9,9(-9,56)/V32/ -9,4(-8,92)//47V -9,5(-8,62)/L76l/ -9,7(-9,24)/V824 

Nelfinavir -10,3(-9,46) -10(-9,44)/D30N -10,1(-9,8)//_90M 

Saquinavir - 1 0,9(- 1 0,46) -1 0,9(-l Q,6)/G48V -1 0,4(-1 0 r 2)/L90M 

Ti pra navi r -1 0,6(- 1 0, 1 ) - 1 0,4(-9,8)//-W - 1 0,3 (- 1 0,04)/O58£ - 1 0,2 (-1 0,0)/T74P -1 0,4{-9,9)/V82L - 1 0,2 

{-9,6)/N83D -1 0,6{-1 0,1 2)/l84V 

Ritonavir -8,0(-7,85) -7,82(-7,44)//471/ -7,8(-7,56)//50/ -7,4{-6,94)/V82L -7,87(-7,58)//84l/ -7.9 

{-7,72)/L90M 



-8,0(-7,66) -8,8{-8,38)/-8,7(-8,28) /-8,8(-8,44)/ -8,8(-8,3)/ -8,8(-8,48)/ -8,7(-8,28)/ -8,8 

(-8,32)/-8,8(-8,48) 

-7,1 (-6,86) -7,9(-7,82)/ -8,2(-8,1 4)/ -8(-7,62)/-8,1 (-7,96)/ -8,2(-8,04)/-8,2(-8,06)/-9(-8,8)/- 

8,2{-8) 

-8,1 (-7,9) -9,4(-9)/ -8,8(-8,56)/-9,4(-8,9)/-9,4(-8,92)/ -9,4(-8,92)/-9,4(-8,88) /-9,4(-8,86)/- 

9,4(-8,88) 

-1 0,7 (-1 0,26) -1 0,4(-1 0,1 8)/- 1 0,1 (-9,92)/- 1 0,8(-1 0,48)/- 1 0,4(-1 0)/ -1 0,4(-1 0,1 8)/-1 0,4(-9,8)/- 

10,4(-10,02)/-10,7(-10,46) 

-9,7(-9,1 4) -9,9(-9,56) A9,7(-9,1 8) /-9,6(-9,24)/-9,3(-8,86)/ -9(-8,32)/-9,6(-9)/-9,5(-9,2)/-9,6 

(-9,42) 

-9,9(-9,46) -10,1 (-9,62) I A 0,4(-9,8)/- 10,1 (-9,74)/- 1 0(-9,84)/ -1 0,1 (-9,72)/- 1 0,1 (-9,74) /-10 

(-9,78)/-10,1(-9,7) 

-10,9 (-10,54) -9,9(-9,64)/-10,3(-9,62) /- 1 0,4(- 1 0,06)/- 1 0,6(- 1 0,44)/ -1 0,6(-1 0,42)/- 1 0,6 

(-1 0,42)/ - 1 0,6(- 1 0,5)/- 1 0,5 (- 1 0,2) 

-9,9(-9,64) -10,2(-9,62) A1 0,1 (-9,92) /-10,3(-9,98)/-10,2(-9,88)/ - 1 0,2(-9,72)/-1 0,3 (-9,72)/- 

1 0,3(-9,72)/-1 0,2(-9,78) 

-7,26(-6,69) -8,3(-7,93)/-7,85(-7,54)/ -7,75(-7,52)/-7,9(-7,76)/ -8,3(-8,1 4)/-8,2(-8,09)/-8,4 

(-8,8)/-7,79(-7,52) 



The average value of the five conformations with less free energy of binding of the protease-PI complex is presented in parentheses. Values that do not correspond to decreasing order of free energy of binding are 
presented in boldface type. 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 12 of 17 



WT protease 



Resistant protease 



D29V mutant protease 







1 


D29V ( / 


f r 






v 


D29V ^* 


tt 



Figure 6 HIV-1 protease structures. Wild-type (WT), resistant and D29V mutant proteases coupled to darunavir (top) and tipranavir (bottom). 
The numbers in the left upper corner are the PDB ID numbers used to model darunavir and tipranavir with the protease and to measure the dis- 
tance (A) between functional groups (purple). 



and TPV (Figure 6). For each natural polymorphism and 
unusual mutation, Table 5 shows the degree of resistance 
to Pis based on free energy of binding differences when 
compared with reference PRs. 

Of the emerging mutations, D29V appears to favour re- 
sistance in silico in seven of nine Pis. Designing more ef- 
fective DRV analogues requires an interaction between D29 
and the bis-tetrahydrofuran ring, as this contributes to 
complex stability [5,42]. All complexes that formed among 
the PRs with natural polymorphisms, unusual mutations 
and drug resistance mutations to TPV and LPV had similar 
free energies of binding. TPV mainly forms hydrogen 
bonds with residues D25, D29, D30, G48 and 150, while 
LPV interacts with G27, D29 and D30. A study that eluci- 
dated the mechanism by which Pis minimize the harmful 
effects of resistance mutations, showed that TPV, ATV, 
LPV, APV, IDV and DRV conserve their antiretroviral 
activity in the presence of drug resistance mutations. This 
phenomenon is due to the compensation of the loss of 
enthalpy (AH) with an entropy gain (-TAS), except in the 
case of TPV [75]. Our results are consistent with another 
report that showed isolated strains with a high level of 
phenotypic resistance to LPV were susceptible to other 
Pis [76]. This corresponds with PR resistance to TPV 
and LPV that contain emerging mutations whose free ener- 
gies of binding were greater than those obtained with wild- 
type PR. 

We found a high prevalence (89%) of L63PGHRS muta- 
tions in HIV-1 variants isolated from Mexican individuals, 
probably because of the prevalence of HIV-1 subtype B 
[32,33]. In the present study, among the functional groups 
found at position 63 (L63G, L63S, L63H and L63R), only 
glycine had hydrophobic characteristics, while serine was 
hydrophilic, and histidine and arginine were alkaline. These 
four mutations conferred resistance to NFV, ATV, TPV and 
LPV, most probably through an allosteric effect, given that 
the substitutions were not located close to the residues 



where the Pis bind [10,74]. Few mutations at position 63 
have been examined for their resistance effects to Pis. The 
L63P mutation has a compensatory effect that increases 
catalytic activity from 110% to 530%; when L63P is associ- 
ated with M46I, it forms a combination that is resistant to 
APV, IDV, LPV or NFV [9,77]. Residue 63 provides hydro- 
phobic contacts between the slit of the loop formed by 
amino acids 38-42 and a |3-sheet (residues 59-63) [74]. Al- 
though the study of mutations in this position has been 
limited to L63P to assess the effect of mutations that pro- 
vide non-hydrophobic characteristics, alternative mecha- 
nisms could be shown by which HIV-1 PR compensates for 
pharmacological stressors. 

Clinical characteristics of patients with unusual mutations 
at resistance sites and/or natural polymorphisms 

Of the participating individuals, 48 of 151 (31.8%) 
showed resistance to at least one PI. Of these, 34 (70.8%) 
showed a high level of resistance, four (8.3%) showed 
intermediate levels of resistance, and 10 (20.8%) showed 
low level resistance. 

Of the 151 sequences, 24 (15.9%) had one or more 
unusual mutations at resistance sites and/or natural poly- 
morphisms. They were isolated from 21 (87.5%) male and 
three (12.5%) female patients; 70.8% of whom lived in the 
central-east of Mexico, and 29.2% in the north-west. Of 
these 24 patients, 23 (95.8%) received antiretroviral therapy, 
and one (4.2%) was naive to treatment. The nucleoside re- 
verse transcriptase inhibitors (NRTIs) and non-nucleoside 
reverse transcriptase inhibitors (NNRTIs) used, in order of 
frequency, were AZT, 3TC, ddl, d4T, ddC, NVP, EFV and 
ABC. The main Pis used were IDV, RTV and SQV. The 
average viral load in this group of patients was 228,225 
virus copies/mL and a mean CD4 + lymphocyte count of 
223 cells/uL. According to the case definition of HIV infec- 
tion and AIDS by the Centers for Disease Control and Pre- 
vention (Atlanta, USA) [78], patients were classified as 



Table 5 Distances (A) between the amino acid of the protease and the PI heteroatoms 







Protease 




aa/ 




Protease 






aa/Chain-PIs 


Wild type 


Resistant 


Emergent 


Chain- 
Pis 


Wild type 


Resistant 


Emergent 


lungui'a ef al. Bi 
^ww.biomedcei 


PR-IDV 


[2R5P] 


[1K6C] 


D29V 


PR-LPV 


[2Q5K] 


[1RV7] 


D29V 


29/A-N 2 


6.1 


6.1 


5.4 


25/A-O4 


5.4 


6.8 


9.1 


29/B-O4 


3.9 


4.2 


4.8 


25/B-O4 


5.5 


7.8 


9.9 


3 n 


30/A-N 2 


6.5 


6.6 


6.3 


29/A-O, 


5.8 


10.8 


6 


Bioinformatics 2C 
l.com/1471-2105 


3O/B-O4 
82/A-0 2 


6.3 
10.1 


6.4 
10.7 


/ 

11.2 


29/B-O3 
30/A-O, 


6 

3.7 


10.3 
9.8 


5.3 
7.1 


82/B-N, 


6.1 


6.2 


7.5 


30/B-O3 


6.7 


10.1 


5.3 


84/A-0 2 


8.3 


8.2 


9.3 


84/A-0 4 


8.2 


10.1 


10.4 


£ > 


84/B-N, 


7.9 


7.9 


8.4 


84/B-O3 


7.6 


8.1 


8.2 


ro in 


PR-SQV 


[30XC] 


[3CYW] 


D29V 


PR-RTV 


[3NDX] 


[1RL8] 


D29V 


NJ 


29/A-0 D , 


4.1 


5.3 


3.5 


29/A-0 /6 


4 


3.8 


4.9 




29/B-N3 


6.1 


6 


7.5 


29/B-Ns 


5.6 


4 


5.1 




30/A-O D , 


3.8 


4.2 


3.8 


30/A-O /6 


6.1 


6.1 


6.9 




3O/B-N3 


6.5 


6.4 


7.5 


3O/B-S3 


4 


5.5 


8.8 




48/A-N D2 


3.4 


5.6 


7.9 


82/A-Nn 


9.4 


9.5 


9.6 




48/B-N 3 


4 


6.3 


6.3 


82/B-0 7 


10.5 


10.3 


11.6 




PR-APV 


[3NU3] 


[3NU4] 


D29V 


PR-DRV 


[IT3R] 


[2HS1] 


D29V 




29/A-0 6 


4.5 


4.2 


4 


29/A-0 26 


3.9 


4.1 


10.4 




29/B-N3 


4.3 


4.6 


7.6 


29/B-N, 


4.6 


4.4 


7.3 




30/A-O 6 


4.5 


3.8 


3.8 


30/A-O 26 


3.8 


3.8 


12.3 




3O/B-N3 


3.3 


3.7 


10.7 


30/B-N, 


3.8 


3.7 


9.5 




32/A-0 6 


6.6 


6.6 


6.9 


32/A-0 26 


7.2 


7.5 


13.9 




32/B-N3 


6.3 


6.3 


15 


32/B-N, 


6.1 


7.6 


13.3 




PR-TPV 


[204P] 


[1DS4] 


D29V 


PR-ATV 


[3EKY] 


[30XX] 


D29V 




29/A-N 2S 


6.1 


6.2 


9.6 


29/A-0 A , 


3.9 


5.1 


5.6 




29/B-O, 


9.1 


9.4 


8.4 


29/B-Oaj 


3.9 


4 


8.3 




30/A-N 2S 


6.2 


6 


8.7 


30/A-O A | 


6.1 


6 


8.7 




30/B-O, 


10 


10.2 


8.6 


30/B-Oaj 


6 


6 


8.5 




82/A-0 8 


10.7 


10 


10.2 


50/A-O, 


5.7 


5.6 


4.8 


3 age 1 3 of 1 7 


82/B-O, 


10.3 


9.9 


/./ 


50/B-O, 


5.3 


5.4 


3.6 



Table 5 Distances (A) between the amino acid of the protease and the PI heteroatoms (Continued) 



84/A-0 8 


8.3 


8.1 


8.5 


84/B-O s 


8.4 


7.8 


8.1 


PR-NFV 


[3EKX] 


[2PYM] 


D29V 


29/A-N 12 


5.9 


6.3 


5.8 


29/B-0 38 


5 


4.4 


3.8 


30/A-N 12 


6.4 


6.7 


6.3 


30/B-O 38 


3.9 


3.7 


3.7 



The classification of atoms corresponds to the PDB file visualized in PyMOL. Measurements correspond to the distance between the alpha carbon of the PR amino acids and the heteroatoms of the inhibitor, and are 
expressed as number of amino acids/chain PR - PI heteroatom. 

Pis, protease inhibitors; aa, amino acid; PR, protease; APV, amprenavir; ATV, atazanavir; DRV, darunavir; IDV, indinavir; LPV, lopinavir; RTV, ritonavir; SQV, saquinavir; TPV, tipranavir. 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 15 of 17 



asymptomatic (n = 2, 8.5%), symptomatic (n = 6, 25%), 
AIDS (« = 12, 50%), and of unknown clinical category (n = 
4, 16.5%). 

Conclusions 

The use of bioinformatics to identify potential mutations that 
confer resistance to antiretroviral drugs allows researchers to 
develop realistic three-dimensional models that illustrate the 
atomic interactions between an enzyme and its substrate. In 
silico, the structural correlation of natural polymorphisms 
and unusual mutations of drug resistance codons, allows the 
identification of HIV-1 variants resistant to Pis. The D29V 
mutation increases the probability of resistance to Pis as it 
generates unstable complexes at the HIV-1 protease active 
site. The prevalence of this mutation in different populations 
should be further studied, and parallel crystallographic stud- 
ies are required to confirm our in silico findings. 

Among mutant PRs-PIs complexes evaluated, TPV and 
LPV had free energies of binding greater than those ob- 
tained with wild-type PRs. 

Furthermore, the presence of a high rate of L63P, I93L, 
V77I and 162V polymorphisms among the Mexican popula- 
tion is similar to that observed in patients that underwent 
antiretroviral treatments in other American and western 
European countries. These data reinforced the knowledge 
regarding the molecular epidemiology of the HIV-1 subtype 
B in Mexico through the presence of HIV polymorphisms. 

Endnote 

The Contents of this publication are the authors re- 
sponsibility and do not necessarily represent the official 
views of the Instituto Mexicano del Seguro Social. 

Additional file 



Additional file 1: Table SI. Identity value, expected value and QMEAN 
analyses of mutant proteases models tested to estimate the quality of 
the predicted structure. 



Abbreviations 

FDA: USA Food and Drug Administration; HAART: Highly Active Antiretrovira 
Therapy; HIV: Human Immunodeficiency Virus; LES: Local Elementary 
Structures; MEGA: Molecular Evolution Genetics Analysis; PDB: Protein Date 
Bank; Pis: Protease Inhibitors; PV: Phenotypic Variation; QMEAN: Qualitative 
Model Energy Analysis; PR: protease; LANL: Los Alamos HIV sequence 
Database; HIVDB: HIV Drug Resistance Database; p: prevalence; M: Mutation; 
ML: Maximun Likelihood; dS: synonimous site divergence; 
dN: nonsynonimous site divergence. 

Competing interests 

The authors declare that they have no competing interest. 
Authors' contributions 

CMM, MED, MVT and MFS performed sequences and bioinformatics analyses. 
EW designed the study. AVO, FGG, MCM and LGFR provided clinical samples 
and collected the data. BTM performed the data analysis, CMM and EW 
wrote the manuscript. All authors read and approved the final manuscript. 



Acknowledgements 

The study was funded in part by the Instituto Mexicano del Seguro Social, 
Grant Number FIS/IMSS/G09/752. The authors wish to thank Ms. Luz Segovia- 
Santos for her editorial services and to Greg Davies for his English revision 
services. 

Author details 

^octorado en Farmacologia, Departamento de Fisiologia, Centra 
Universitario de Ciencias de la Salud, Universidad de Guadalajara, Guadalajara 
44340, Mexico. 2 Laboratorio de Inmunodeficiencias y Retrovirus Humanos, 
Centra de Investigation Biomedica de Occidente, CMNO, IMSS, Guadalajara 
44340, Mexico. 3 Unidad de Investigation Cardiovascular, Departamento de 
Fisiologia, Centra Universitario de Ciencias de la Salud, Universidad de 
Guadalajara, Guadalajara 44340, Mexico. 4 Divisi6n de Excelencia Clinica, 
Coordination Medica de Unidades de Alta Especialidad, Unidad de Atencion 
Medica, IMSS, Mexico, D.F 06700, Mexico. 5 UMAE, Hospital de Especialidades, 
CMNO, IMSS, Guadalajara 44340, Mexico. 6 Departamento de Production 
Agricola, Centra Universitario de Ciencias Biologicas y Agropecuarias, 
Universidad de Guadalajara, Zapopan 45110, Mexico. 7 Departamento de 
Clinicas Medicas, Centra Universitario de Ciencias de la Salud, Universidad de 
Guadalajara, Guadalajara 44340, Mexico. 8 Departamento de Farmacobiologia, 
CUCEI, Universidad de Guadalajara, Guadalajara 44430, Mexico. 

Received: 27 June 2012 Accepted: 5 March 2014 
Published: 15 March 2014 



References 

1. Shafer RW, Kantor R, Gonzales MJ: The genetic basis of HIV-1 resistance to 
reverse transcriptase and protease inhibitors. AIDS Rev 2000, 2:21 1-228. 

2. Rambaut A, Posada D, Crandall KA, Holmes EC: The causes and 
consequences of HIV evolution. Nat Rev Genet 2004, 5:52-61. 

3. Ribeiro RM, Bonhoeffer S: Production of resistant HIV mutants during 
antiretroviral therapy. Proc Natl Acad Sci USA 2000, 97:7681 -7686. 

4. Adamson CS, Freed EO: Recent progress in antiretrovirals - lessons from 
resistance. Drug Discov Today 2008, 13:424-432. 

5. Ghosh AK, Sridhar PR, Leshchenko S, Hussain AK, Li J, Kovalevsky AY, Walters 
DE, Wedekind JE, Grum-Tokars V, Das D, Koh Y, Maeda K, Gatanaga H, 
Weber IT, Mitsuya H: Structure-based design of novel HIV-1 protease 
inhibitors to combat drug resistance. J Med Chem 2006, 49:5252-5261. 

6. Ohtaka H, Schon A, Freire E: Multidrug resistance to HIV-1 protease inhib- 
ition requires cooperative coupling between distal mutations. 
Biochemistry 2003, 42:1 3659-1 3666. 

7. Broglia RA, Provasi D, Vasile F, Ottolina G, Longhi R, Tiana G: A folding 
inhibitor of the HIV-1 protease. Proteins 2006, 62:928-933. 

8. Broglia RA, Tiana G, Sutto L, Provasi D, Simona F: Design of HIV-1-PR inhibi- 
tors that do not create resistance: blocking the folding of single mono- 
mers. Protein Sci 2005, 14:2668-2681. 

9. Piana S, Carloni P, Rothlisberger U: Drug resistance in HIV-1 protease: 
Flexibility-assisted mechanism of compensatory mutations. Protein Sci 
2002, 11:2393-2402. 

10. Hornak V, Simmerling C: Targeting structural flexibility in HIV-1 protease 
inhibitor binding. Drug Discov Today 2007, 12:132-138. 

1 1 . Shafer RW, Rhee SY, Pillay D, Miller V, Sandstrom P, Schapiro JM, Kuritzkes 
DR, Bennett D: HIV-1 protease and reverse transcriptase mutations for 
drug resistance surveillance. AIDS 2007, 21:215-223. 

12. Johnson VA, Calvez V, Gunthard HF, Paredes R, Pillay D, Shafer R, Wensing 
AM, Richman DD: Update of the drug resistance mutations in HIV-1: 
March 2013. Top Antivir Med 2013, 21:4-12. 

13. Johnson JA, Li JF, Wei X, Lipscomb J, Irlbeck D, Craig C, Smith A, Bennett 
DE, Monsour M, Sandstrom P, Lanier ER, Heneine W: Minority HIV-1 drug 
resistance mutations are present in antiretroviral treatment-naive popu- 
lations and associate with reduced treatment efficacy. PLoS Med 2008, 
5:el58. 

14. Benson DA, Karsch-Mizrachi I, Lipman DJ, Ostell J, Rapp BA, Wheeler DL: 
GenBank. Nucleic Acids Res 2002, 30:17-20. 

15. Rose PP, Korber BT: Detecting hypermutations in viral sequences with an 
emphasis on G - > A hypermutation. Bioinformatics 2000, 16:400-401. 

16. Rhee SY, Gonzales MJ, Kantor R, Betts BJ, Ravela J, Shafer RW: Human 
immunodeficiency virus reverse transcriptase and protease sequence 
database. Nucleic Acids Res 2003, 31:298-303. 



Mata-Mungui'a et al. BMC Bioinformatics 2014, 15:72 
http://www.biomedcentral.com/1471-2105/15/72 



Page 16 of 17 



1 7. Rozanov M, Plikat U, Chappey C, Kochergin A, Tatusova T: A web-based 
genotyping resource for viral sequences. Nucleic Acids Res 2004, 
32:W654-W659. 

18. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S: MEGA5: 
Molecular evolutionary genetics analysis using maximum likelihood, 
evolutionary distance, and maximum parsimony methods. Mol Biol Evol 
2011,28:2731-2739. 

1 9. Kuiken C, Korber B, Shafer RW: HIV sequence databases. AIDS Rev 2003, 
5:52-61. 

20. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam 
H, Valentin F, Wallace IM, Wilm A, Lopez R: Clustal W and Clustal X version 
2.0. Bioinformatics 2007, 23:2947-2948. 

21. Kimura (VI: A simple method for estimating evolutionary rates of base 
substitutions through comparative studies of nucleotide sequences. 
J Mol Evol 1980, 16:111-120. 

22. Rhee SY, Kantor R, Katzenstein DA, Camacho R, Morris L, Sirivichayakul S, 
Jorgensen L, Brigido LF, Schapiro JM, Shafer RW, for the International Non 
Subtype: HIV-1 pol mutation frequency by subtype and treatment 
experience: extension of the HIVseq program to seven non-B subtypes. 
ADS 2006, 20:643-651. 

23. Pond SL, Frost SD, Muse SV: HyPhy: hypothesis testing using phylogenies. 
Bioinformatics 2005, 21:676-679. 

24. Yang Z, Goldman N, Friday A: Comparison of models for nucleotide 
substitution used in maximum-likelihood phylogenetic estimation. 
Mol Biol EvoH 994, 11:316-324. 

25. Muse SV, Gaut BS: A likelihood approach for comparing synonymous and 
nonsynonymous nucleotide substitution rates, with application to the 
chloroplast genome. Mol Biol Evol 1 994, 1 1 :71 5-724. 

26. Pan C, Kim J, Chen L, Wang Q, Lee C: The HIV positive selection mutation 
database. Nucleic Acids Res 2007, 35:D371-D375. 

27. Bordoli L, Kiefer F, Arnold K, Benkert P, Battey J, Schwede T: Protein 
structure homology modeling using SWISS-MODEL workspace. 
Nat Protocols 2008, 4:1-13. 

28. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, 
Shindyalov IN, Bourne PE: The protein data bank. Nucleic Acids Res 2000, 
28:235-242. 

29. Benkert P, Tosatto SCE, Schomburg D: QMEAN: a comprehensive scoring 
function for model quality assessment. Proteins 2008, 71:261-277. 

30. Trott 0, Olson AJ: AutoDock Vina: improving the speed and accuracy of 
docking with a new scoring function, efficient optimization, and 
multithreading. J Comput Chem 2010, 31:455-461. 

31. Delano WL: The PyMOL Molecular Graphics System. 2002 [http://www.pymol. 
org] 

32. Vazquez-Vails E, Escoto-Delgadillo M, Lopez-Marquez FC, Castillero-Manzano 
M, Echegaray-Guerrero E, Bitzer-Quintero OK, Kobayashi-Gutierrez A, Torres- 
Mendoza BM: Molecular epidemiology of HIV type 1 in Mexico: emer- 
gence of BG and BF intersubtype recombinants. AIDS Res Hum Retroviruses 
2010, 26:777-781. 

33. Avila-Rios S, Garcia-Morales C, Garrido-Rodriguez D, Ormsby CE, Hernandez- 
Juan R, Andrade-Villanueva J, Gonzalez-Hernandez LA, Torres-Escobar I, 
Navarro-Alvarez S, Reyes-Teran G, Mexican HIV Molecular Epidemiology Pro- 
ject Group: National prevalence and trends of HIV transmitted drug re- 
sistance in Mexico. PLoS One 201 1, 6:e27812. 

34. Hartman JL, Garvik B, Hartwell L: Principles for the buffering of genetic 
variation. Science 2001, 9(291):1001-4. 

35. Burger R, Willensdorfer M, Nowak MA: Why are phenotypic mutation rates 
much higher than genotypic mutation rates? Genetics 2006, 172:197-206. 

36. Rhee SY, Liu TF, Holmes SP, Shafer RW: HIV-1 subtype b protease and 
reverse transcriptase amino acid covariation. PLoS Comput Biol 2007, 
3:e87. 

37. Chen L, Lee C: Distinguishing HIV-1 drug resistance, accessory, and viral 
fitness mutations using conditional selection pressure analysis of treated 
versus untreated patient samples. Biol Direct 2006, 1:14. 

38. Bonomi M, Gervasio FL, Tiana G, Provasi D, Broglia RA, Parrinello M: Insight 
into the folding inhibition of the HIV-1 protease by a small peptide. 
Biophys J 2007, 93:2813-2821. 

39. Condra JH, Holder DJ, Schleif WA, Blahy OM, Danovich RM, Gabryelski U, 
Graham DJ, Laird D, Quintero JC, Rhodes A, Robbins HL, Roth E, 
Shivaprakash M, Yang T, Chodakewitz JA, Deutsch PJ, Leavitt RY, Massari FE, 
Mellors JW, Squires KE, Steigbigel RT, Teppler H, Emini EA: Genetic 
correlates of in vivo viral resistance to indinavir, a human 



immunodeficiency virus type 1 protease inhibitor. J Virol 1996, 
70:8270-8276. 

40. Zhang J, Hou T, Wang W, Liu JS: Detecting and understanding 
combinatorial mutation patterns responsible for HIV drug resistance. 
Proc Natl Acad Sci USA 2010, 107:1321-1326. 

41 . Shafer RW, Rhee SY, Bennett DE: Consensus drug resistance mutations for 
epidemiological surveillance: basic principles and potential 
controversies. Antivir Jher 2008, 13(Suppl 2):59-68. 

42. Ghosh AK, Chapsal BD, Weber IT, Mitsuya H: Design of HIV protease 
inhibitors targeting protein backbone: an effective strategy for 
combating drug resistance. Acc Chem Res 2007, 41:78-86. 

43. Haq O, Levy R, Morozov A, Andrec M: Pairwise and higher-order correla- 
tions among drug-resistance mutations in HIV-1 subtype B protease. 
BMC Bioinformatics 2009, 10:510. 

44. Melnick L, Yang SS, Rossi R, Zepp C, Heefner D: An Escherichia coli 
expression assay and screen for human immunodeficiency virus 
protease variants with decreased susceptibility to indinavir. 
Antimicrob Agents Chemother 1998, 42:3256-65. 

45. Svicher V, Ceccherini-Silberstein F, Erba F, Santoro M, Gori C, Bellocchi MC, 
Giannella S, Trotta MP, Monforte A, Antinori A, Perno CF: Novel human 
immunodeficiency virus type 1 protease mutations potentially involved 
in resistance to protease inhibitors. Antimicrob Agents Chemother 2005, 
49:2015-2025. 

46. Jacobsen H, Yasargil K, Winslow DL, Craig JC, Krohn A, Duncan IB, Mous J: 
Characterization of human immunodeficiency virus type 1 mutants with 
decreased sensitivity to proteinase inhibitor Ro 31-8959. Virology 1995, 
206:527-534. 

47. Markowitz M, Conant M, Hurley A, Schluger R, Duran M, Peterkin J, 
Chapman S, Patick A, Hendricks A, Yuen GJ, Hoskins W, Clendeninn N, Ho 
DD: A preliminary evaluation of nelfinavir mesylate, an inhibitor of 
human immunodeficiency virus (HIV)-I protease, to treat HIV infection. 
J Infect Dis 1998, 177:1533-40. 

48. Marcelin AG, Flandre P, De Mendoza C, Roquebert B, Peytavin G, Valer L, 
Wirden M, Abbas S, Katlama C, Soriano V, Calvez V: Clinical validation of 
saquinavir/ritonavir genotypic resistance score in protease-inhibitor- 
experienced patients. Antivir Jher 2007, 1 2:247-52. 

49. Marcelin AG, Lamotte C, Delaugerre C, Ktorza N, Ait Mohand H, Cacace R, 
Bonmarchand M, Wirden M, Simon A, Bossi P, Bricaire F, Costagliola D, 
Katlama C, Peytavin G, Calvez V, Genophar Study Group: Genotypic 
inhibitory quotient as predictor of virological response to ritonavir- 
amprenavir in human immunodeficiency virus type 1 protease inhibitor- 
experienced patients. Antimicrob Agents Chemother 2003, 47:594-600. 

50. Molla A, Brun S, Mo H, Real K, Podding J, Bernstein B, Hertogs K, Larder B, 
Lie Y, Hellmann N, Vasavanonda S, Chernyavskiy T, Freimuth W, Japour A, 
Sun E, Kempf D: Genotypic and phenotypic analysis of viral isolates from 
subjects with detectable viral load on therapy with ABT-378/ritonavir. 
Antiviral Ther 2000, 5(Suppl 3)30 

51 . Harrigan PR, Hertogs K, Verbiest W, Pauwels R, Larder B, Kemp S, Bloor S, Yip 
B, Hogg R, Alexander C, Montaner JS: Baseline HIV drug resistance profile 
predicts response to ritonavir-saquinavir protease inhibitor therapy in a 
community setting. AIDS 1999, 13:1863-1871. 

52. Shafer RW, Hsu P, Patick AK, Craig C, Brendel V: Identification of biased 
amino acid substitution patterns in human immunodeficiency virus type 
1 isolates from patients treated with protease inhibitors. J Virol 1999, 
73:6197-6202. 

53. Carrillo A, Stewart KD, Sham HL, Norbeck DW, Kohlbrenner WE, Leonard JM, 
Kempf DJ, Molla A: In vitro selection and characterization of human 
immunodeficiency virus type 1 variants with increased resistance to 
ABT-378, a novel protease inhibitor. J Virol 1998, 72:7532-7541. 

54. De Meyer S, Azijn H, Van Ginderen M, De Baere I, Pauwels R, De Bethune 
MP: In vitro selection experiments demonstrate an increased genetic 
barrier to resistance development to TMC1 14 as compared with 
currently licensed protease inhibitors. Antiviral Ther 2002, 7:S5. 

55. Coman RM, Robbins AH, Fernandez MA, Gilliland CT, Sochet AA, Goodenow 
MM, McKenna R, Dunn BM: The contribution of naturally occurring 
polymorphisms in altering the biochemical and structural characteristics 
of HIV-1 subtype C protease. Biochemistry 2008, 47:731-743. 

56. Kantor R, Katzenstein DA, Efron B, Carvalho AP, Wynhoven B, Cane P, Clarke 
J, Sirivichayakul S, Soares MA, Snoeck J, Pillay C, Rudich H, Rodrigues R, 
Holguin A, Ariyoshi K, Bouzas MB, Cahn P, Sugiura W, Soriano V, Brigido LF, 
Grossman Z, Morris L, Vandamme AM, Tanuri A, Phanuphak P, Weber JN, 



Mata-Mungui'a ef al. BMC Bioinformatics 2014, 15:72 Page 17 of 17 

http://www.biomedcentral.com/1471-2105/15/72 



57. 



58. 



50 



Pillay D, Harrigan PR, Camacho R, Schapiro JM, et al. Impact of HIV-1 sub- 
type and antiretroviral therapy on protease and reverse transcriptase 
genotype: results of a global collaboration. PLoS Med 2005, 2:el 1 2. 
The HIVdb Release Notes - HIV Drug Resistance Database [http://hivdb. 
stanford.edu/DR/asi/releaseNotes/index.html] 

Ali A, Bandaranayake RM, Cai Y, King NM, Kolli M, Mittal S, Murzycki JF, 
Nalam MN, Nalivaika EA, Ozen A, Prabu-Jeyabalan MM, Thayer K, Schiffer CA: 
Molecular basis for drug resistance in HIV-1 protease. Viruses 2010, 
2:2509-2535. 

Kozisek M, Bray J, Rezacova P, Saskova K, Brynda J, Pokorna J, Mammano F, 
Rulisek L, Konvalinka J: Molecular analysis of the HIV-1 resistance develop- 
ment: enzymatic activities, crystal structures, and thermodynamics of 
nelfinavir-resistant HIV protease mutants. J Mol Biol 2007, 374:1005-1016. 

60. Broglia RA: Learning to design resistance proof drugs from folding. 
Eur Phys J D At Mol Opt Plasma Phys 2009, 51:137-151. 

61. Parera M, Fernandez G, Clotet B, Martjnez MA: HIV-1 protease catalytic 
efficiency effects caused by random single amino acid substitutions. 
Mol Biol Evol 2007, 24:382-387. 

Ceccherini-Silberstein F, Erba F, Gago F, Bertoli A, Forbici F, Bellocchi MC, 
Gori C, D'Arrigo R, Marcon L, Balotta C, Antinori A, Monforte A, Perno CF: 
Identification of the minimal conserved structure of HIV-1 protease in 
the presence and absence of drug pressure. AIDS 2004, 18:F1 1-F19. 
Boeri E, Gianotti N, Canducci F, Hasson H, Giudici B, Castagna A, Lazzarin A, 
Clementi M: Evolutionary characteristics of HIV type 1 variants resistant 
to protease inhibitors in the absence of drug-selective pressure. AIDS Res 
Hum Retroviruses 2003, 19:1 151-1 153. 

Karlsson AC, Deeks SG, Barbour JD, Heiken BD, Younger SR, Hoh R, Lane (VI, 
Sallberg M, Ortiz GM, Demarest JF, Liegler T, Grant RM, Martin JN, Nixon D: 
Dual pressure from antiretroviral therapy and cell-mediated immune 
response on the human immunodeficiency virus type 1 protease gene. 

J Virol 2003, 77:6743-6752. 

Prabu-Jeyabalan M, Nalivaika EA, King NM, Schiffer CA: Viability of a drug- 
resistant human immunodeficiency virus type 1 protease variant: struc- 
tural insights for better antiviral therapy. J Virol 2003, 77:1306-1315. 
De Oliveira T, Salemi M, Gordon M, Vandamme AM, Van Rensburg EJ, 
Engelbrecht S, Coovadia HM, Cassol S: Mapping sites of positive selection 
and amino acid diversification in the HIV genome. Genetics 2004, 
167:1047-1058. 

Arnold K, Bordoli L, Kopp J+, Schwede T: The SWISS-MODEL workspace: 
a web-based environment for protein structure homology modelling. 

Bioinformatics 2006, 22:1 95-201 . 

Bonomi M, Barducci A, Gervasio FL, Parrinello M: Multiple routes and 
milestones in the folding of HIV-1 protease monomer. PLoS ONE 2010, 
5:el3208. 

Levy Y, Caflisch A, Onuchic JN, Wolynes PG: The folding and dimerization 
of HIV-1 protease: evidence for a stable monomer from simulations. 

J Mol Biol 2004, 340:67-79. 

Sharp PM, Hahn BH: The evolution of HIV-1 and the origin of AIDS. 

Philos Trans R Soc Lond B Biol Sci 2010, 365:2487-2494. 
Wu TD, Schiffer CA, Gonzales MJ, Taylor J, Kantor R, Chou S, Israelski D, 
Zolopa AR, Fessel WJ, Shafer RW: Mutation patterns and structural 
correlates in human immunodeficiency virus type 1 protease following 
different protease inhibitor treatments. J Virol 2003, 77:4836-4847. 
Bromberg Y, Rost B: Correlating protein function and stability through 
the analysis of single amino acid substitutions. BMC Bioinformatics 2009, 
10:S8. 

Gao C, Park MS, Stern HA: Accounting for ligand conformational 
restriction in calculations of protein-ligand binding affinities. Biophys J 
2010, 98:901-910. 

Heaslet H, Rosenfeld R, Giffin M, Lin YC, Tam K, Torbett BE, Elder JH, McRee 
DE, Stout CD: Conformational flexibility in the flap domains of ligand-free 
HIV protease. Acta Crystallogr D Biol Crystallogr 2007, 63:866-875. 
Muzammil S, Armstrong AA, Kang LW, Jakalian A, Bonneau PR, Schmelmer 
V, Amzel LM, Freire E: Unique thermodynamic response of tipranavir to 
human immunodeficiency virus type 1 protease drug resistance 
mutations. J Virol 2007, 81:5144-5154. 

Friend J, Parkin N, Liegler T, Martin JN, Deeks SG: Isolated lopinavir 
resistance after virological rebound of a ritonavir/lopinavir-based 
regimen. AIDS 2004, 18:1965-1966. 

Weikl TR, Hemmateenejad B: How conformational changes can affect 
catalysis, inhibition and drug resistance of enzymes with induced-fit 



63 



64 



65 



66 



67. 



69. 



~0 



72 



73. 



74. 



76. 



binding mechanism such as the HIV-1 protease. Biochim Biophys Acta 
2013, 1834:867-873. 
78. Centers for Disease Control and Prevention: 1993 revised classification 
system for HIV infection and expanded surveillance case definition for 
AIDS among adolescents and adults. MMWR Recomm Rep 1992, 
41(RR-17):1-19. 



doi:1 0.1 1 86/1 471 -21 05-1 5-72 

Cite this article as: Mata-Munguia ef al: Natural polymorphisms and 
unusual mutations in HIV-1 protease with potential antiretroviral 
resistance: a bioinformatic analysis. BMC Bioinformatics 2014 15:72. 



Submit your next manuscript to BioMed Central 
and take full advantage of: 

• Convenient online submission 

• Thorough peer review 

• No space constraints or color figure charges 

• Immediate publication on acceptance 

• Inclusion in PubMed, CAS, Scopus and Google Scholar 

• Research which is freely available for redistribution 



Submit your manuscript at 
www.biomedcentral.com/submit 



o 



BioMed Central